© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 11917
    Prev
    Next
    INDEX
    Explanations

    terms related to bankruptcy and financial processes

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/4-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.4.hook_resid_post
    Hook Layer
    4
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    332
    +0.16
    0.9%
    262
    +0.15
    0.8%
    82
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    332
    +0.16
    0.07
    168
    +0.15
    0.05
    57
    +0.12
    0.05
    Negative Logits
     tumours
    -2.10
     flavour
    -2.02
     labelling
    -2.01
     favourite
    -1.99
     fibres
    -1.93
     cyclists
    -1.86
     colour
    -1.85
     behavioural
    -1.84
     anaest
    -1.79
     armour
    -1.79
    POSITIVE LOGITS
     Debtor
    3.65
     Bankruptcy
    3.53
     debtor
    3.40
     Debtors
    3.13
     Trustee
    3.08
     creditors
    3.01
     bankruptcy
    2.92
     debtors
    2.75
    ruptcy
    2.64
     creditor
    2.60
    Activations Density 0.597%

    No Known Activations